[quant] Rename "2.0" to "2", remove "static" (#2576)

andrewor14 · jerryzh168 · web-flow · commit 5de96a8703a8 · 2023-10-02T11:05:52.000-07:00
* [quant] Rename "2.0" to "2", remove "static"
---------

Co-authored-by: Jerry Zhang &lt;jerryzh168@gmail.com&gt;
diff --git a/prototype_source/prototype_index.rst b/prototype_source/prototype_index.rst
@@ -69,17 +69,17 @@ Prototype features are not available as part of binary distributions like PyPI o
    :tags: Debugging,Quantization
 
 .. customcarditem::
-   :header: How to Write a Quantizer for PyTorch 2.0 Export Quantization
-   :card_description: Learn how to implement a Quantizer for PT2.0 Export Quantization
+   :header: How to Write a Quantizer for PyTorch 2 Export Quantization
+   :card_description: Learn how to implement a Quantizer for PT2 Export Quantization
    :image: ../_static/img/thumbnails/cropped/generic-pytorch-logo.png
    :link: ../prototype/pt2e_quantizer.html
    :tags: Quantization
 
 .. customcarditem::
-   :header: PyTorch 2.0 Export Post Training Static Quantization
-   :card_description: Learn how to use Post Training Static Quantization in PyTorch 2.0 Export.
+   :header: PyTorch 2 Export Post Training Quantization
+   :card_description: Learn how to use Post Training Quantization in PyTorch 2 Export.
    :image: ../_static/img/thumbnails/cropped/generic-pytorch-logo.png
-   :link: ../prototype/pt2e_quant_ptq_static.html
+   :link: ../prototype/pt2e_quant_ptq.html
    :tags: Quantization          
 
 .. customcarditem::
@@ -223,8 +223,10 @@ Prototype features are not available as part of binary distributions like PyPI o
    prototype/fx_graph_mode_ptq_dynamic.html
    prototype/fx_graph_mode_ptq_static.html
    prototype/graph_mode_dynamic_bert_tutorial.html
-   prototype/quantization_in_pytorch_2_0_export_tutorial.html
    prototype/inductor_cpp_wrapper_tutorial.html
+   prototype/pt2e_quantizer.html
+   prototype/pt2e_quant_ptq.html
+   prototype/pt2e_quant_qat.html
    prototype/ios_gpu_workflow.html
    prototype/nnapi_mobilenetv2.html
    prototype/tracing_based_selective_build.html
diff --git a/prototype_source/pt2e_quant_ptq.rst b/prototype_source/pt2e_quant_ptq.rst
@@ -1,4 +1,4 @@
-(prototype) PyTorch 2.0 Export Post Training Static Quantization
+(prototype) PyTorch 2 Export Post Training Quantization
 ================================================================
 **Author**: `Jerry Zhang <https://github.com/jerryzh168>`_
 
@@ -13,7 +13,7 @@ better programmability, and a simplified UX.
 Exportable by `torch.export.export` is a prerequisite to use the flow, you can
 find what are the constructs that's supported in `Export DB <https://pytorch.org/docs/main/generated/exportdb/index.html>`_.
 
-The high level architecture of quantization 2.0 with quantizer could look like
+The high level architecture of quantization 2 with quantizer could look like
 this:
 
 ::
@@ -46,7 +46,7 @@ this:
             Executorch, Inductor or <Other Backends>
 
 
-The PyTorch 2.0 export quantization API looks like this:
+The PyTorch 2 export quantization API looks like this:
 
 .. code:: python
 
@@ -93,10 +93,10 @@ The PyTorch 2.0 export quantization API looks like this:
   # we have a model with aten ops doing integer computations when possible
 
 
-Motivation of PyTorch 2.0 Export Quantization
+Motivation of PyTorch 2 Export Quantization
 ---------------------------------------------
 
-In PyTorch versions prior to 2.0, we have FX Graph Mode Quantization that uses
+In PyTorch versions prior to 2, we have FX Graph Mode Quantization that uses
 `QConfigMapping <https://pytorch.org/docs/main/generated/torch.ao.quantization.qconfig_mapping.QConfigMapping.html>`_
 and `BackendConfig <https://pytorch.org/docs/stable/generated/torch.ao.quantization.backend_config.BackendConfig.html>`_
 for customizations. ``QConfigMapping`` allows modeling users to specify how
@@ -394,7 +394,7 @@ the different configuration APIs supported by ``XNNPackQuantizer``:
    `tutorial <https://pytorch.org/tutorials/prototype/pt2e_quantizer.html>`_
    that describes how to write a new ``Quantizer``.
 
-Prepare the Model for Post Training Static Quantization
+Prepare the Model for Post Training Quantization
 ----------------------------------------------------------
 
 ``prepare_pt2e`` folds ``BatchNorm`` operators into preceding ``Conv2d``
@@ -572,7 +572,7 @@ Debugging the Quantized Model
 
 You can use `Numeric Suite <https://pytorch.org/docs/stable/quantization-accuracy-debugging.html#numerical-debugging-tooling-prototype>`_
 that can help with debugging in eager mode and FX graph mode. The new version of
-Numeric Suite working with PyTorch 2.0 Export models is still in development.
+Numeric Suite working with PyTorch 2 Export models is still in development.
 
 Lowering and Performance Evaluation
 ------------------------------------
@@ -587,7 +587,7 @@ operators.
 Conclusion
 --------------
 
-In this tutorial, we went through the overall quantization flow in PyTorch 2.0
+In this tutorial, we went through the overall quantization flow in PyTorch 2
 Export Quantization using ``XNNPACKQuantizer`` and got a quantized model that
 could be further lowered to a backend that supports inference with XNNPACK
 backend. To use this for your own backend, please first follow the
diff --git a/prototype_source/pt2e_quantizer.rst b/prototype_source/pt2e_quantizer.rst
@@ -1,10 +1,8 @@
-How to Write a ``Quantizer`` for PyTorch 2.0 Export Quantization
+How to Write a ``Quantizer`` for PyTorch 2 Export Quantization
 ================================================================
 
 **Author**: `Leslie Fang <https://github.com/leslie-fang-intel>`_, `Weiwen Xia <https://github.com/Xia-Weiwen>`__, `Jiong Gong <https://github.com/jgong5>`__, `Kimish Patel <https://github.com/kimishpatel>`__, `Jerry Zhang <https://github.com/jerryzh168>`__
 
-.. note:: Quantization in PyTorch 2.0 export is still a work in progress.
-
 Prerequisites:
 ^^^^^^^^^^^^^^^^
 
@@ -14,7 +12,7 @@ Required:
    
 -  `Quantization concepts in PyTorch <https://pytorch.org/docs/master/quantization.html#quantization-api-summary>`__
    
--  `(prototype) PyTorch 2.0 Export Post Training Static Quantization <https://pytorch.org/tutorials/prototype/pt2e_quant_ptq_static.html>`__
+-  `(prototype) PyTorch 2 Export Post Training Quantization <https://pytorch.org/tutorials/prototype/pt2e_quant_ptq.html>`__
 
 Optional:
 
@@ -27,11 +25,11 @@ Optional:
 Introduction
 ^^^^^^^^^^^^^
 
-`(prototype) PyTorch 2.0 Export Post Training Static Quantization <https://pytorch.org/tutorials/prototype/pt2e_quant_ptq_static.html>`__ introduced the overall API for pytorch 2.0 export quantization, main difference from fx graph mode quantization in terms of API is that we made it explicit that quantiation is targeting a specific backend. So to use the new flow, backend need to implement a ``Quantizer`` class that encodes:
+`(prototype) PyTorch 2 Export Post Training Quantization <https://pytorch.org/tutorials/prototype/pt2e_quant_ptq.html>`__ introduced the overall API for pytorch 2 export quantization, main difference from fx graph mode quantization in terms of API is that we made it explicit that quantiation is targeting a specific backend. So to use the new flow, backend need to implement a ``Quantizer`` class that encodes:
 (1). What is supported quantized operator or patterns in the backend
 (2). How can users express the way they want their floating point model to be quantized, for example, quantized the whole model to be int8 symmetric quantization, or quantize only linear layers etc.
 
-Please see `here <https://pytorch.org/tutorials/prototype/pt2e_quant_ptq_static.html#motivation-of-pytorch-2-0-export-quantization>`__ For motivations for the new API and ``Quantizer``.
+Please see `here <https://pytorch.org/tutorials/prototype/pt2e_quant_ptq.html#motivation-of-pytorch-2-export-quantization>`__ For motivations for the new API and ``Quantizer``.
 
 An existing quantizer object defined for ``XNNPACK`` is in
 `QNNPackQuantizer <https://github.com/pytorch/pytorch/blob/main/torch/ao/quantization/pt2e/quantizer/xnnpack_quantizer.py>`__
@@ -307,7 +305,7 @@ functions that are used in the example:
 Conclusion
 ^^^^^^^^^^^^^^^^^^^
 
-With this tutorial, we introduce the new quantization path in PyTorch 2.0. Users can learn about
-how to define a ``BackendQuantizer`` with the ``QuantizationAnnotation API`` and integrate it into the quantization 2.0 flow.
+With this tutorial, we introduce the new quantization path in PyTorch 2. Users can learn about
+how to define a ``BackendQuantizer`` with the ``QuantizationAnnotation API`` and integrate it into the PyTorch 2 Export Quantization flow.
 Examples of ``QuantizationSpec``, ``SharedQuantizationSpec``, ``FixedQParamsQuantizationSpec``, and ``DerivedQuantizationSpec``
-are given for specific annotation use case. This is a prerequisite to be able to quantize a model in PyTorch 2.0 Export Quantization flow. You can use `XNNPACKQuantizer <https://github.com/pytorch/pytorch/blob/main/torch/ao/quantization/quantizer/xnnpack_quantizer.py>`_ as an example to start implementing your own ``Quantizer``. After that please follow `this tutorial <https://pytorch.org/tutorials/prototype/pt2e_quant_ptq_static.html>`_ to actually quantize your model.
+are given for specific annotation use case. You can use `XNNPACKQuantizer <https://github.com/pytorch/pytorch/blob/main/torch/ao/quantization/quantizer/xnnpack_quantizer.py>`_ as an example to start implementing your own ``Quantizer``. After that please follow `this tutorial <https://pytorch.org/tutorials/prototype/pt2e_quant_ptq.html>`_ to actually quantize your model.
